Demo — every number on this page is fake. Synthetic data seeded locally so the design can be reviewed. The five kernel-panic incidents are real (imported from research/panic-log.md); everything else, including all response times, is generated.

robbiesucks.lol

Service status

Loading…

Loading…

Operational Degraded Some downtime Significant downtime Not monitored Contains imported history

Show data tables

Preview state:

Response time

Incident history

How this works

Where it runs

A Cloudflare Worker on a one-minute Cron Trigger, running on Cloudflare's network — not on the machine it watches. A status page served by the box would be blank at exactly the moment it mattered.

What counts as up

Measured vs imported

Anything tagged imported — a bar, a percentage, a stat tile, an incident — was reconstructed from the machine's own kernel-panic log rather than observed by this monitor. It is shown so the history is complete, and tagged so it is never mistaken for measurement.

Why some numbers say “≤”

Availability is downtime against the length of the period, not against the checks that happen to exist. Where part of a period was never measured the honest answer is an upper bound — the unmeasured minutes could only have made it worse — so the page prints ≤ and says how much was missing.

What it can't see

The prober and this page both run on Cloudflare. A Cloudflare-wide failure would take both down together, and this page measures from Cloudflare's network rather than from your ISP — so a last-mile or DNS problem on your side won't appear here.